the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 08 Jan 2025
Ozone decrease observed in the upper atmosphere following the May 11th 2024 Mother’s day solar storm
Abstract. On May 11th 2024, a succession of coronal mass ejections that merged together struck the Earth and induced large scale perturbations in the magnetosphere. During this event, satellite observations showed a large solar energetic proton (SEP) event associated to an extreme geomagnetic storm. At the same time, satellite observations of atmospheric ozone have been performed by AURA/MLS. In this work, we present the first observations of the effect of the storm of May and the following SEP of June 8th on ozone concentration throughout the atmosphere. Observations of the MLS show that the event of May lead to stronger depletion of O3 in the upper part of the atmosphere than in June. This difference is explained by the type of particle precipitation that occurred during the two events, with both protons and electrons in May and only protons in June. Neither event caused ozone depletion in the stratosphere while strong decreases are observed in the mesosphere. In May, mesospheric ozone depletion is observed during 18 days and reaches a maximum of 60 %. In addition, the storm of May also caused a noticeable decrease in ozone concentration (up to 20 %) at altitudes above 90 km.
- Preprint
(2013 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on angeo-2024-29', Anonymous Referee #1, 03 Feb 2025
Review of "Ozone decrease observed in the upper atmosphere following the
May 11th 2024 Mother’s day solar storm" by A. Winant et al.The authors analyse MLS observation in early 2025 for examining the impact of geomagnetic storm on May-11 2024 on ozone in the middle atmosphere and MLT region. They find additional ozone depletion related to the event for the secondary and tertiary ozone layer in the Southern hemisphere. In addition, they study the solar proton event in June 2024 for comparison. None of these events caused long-lived ozone depletion in the stratosphere.
Through its worlwide observable auroras, the geomagnetic storms of May 2024 draw broad public attention. But also from a scientific perspective, the geomagnetic storm in May 2024 is interesting as it exhibited the second strongest Ap-index recorded, and ranks among the strongest storms in terms of the Dst-index. The paper is one of the first publications to study specifically the impact of this storm on the neutral atmosphere. The paper is timely and of general interest.
The paper is generally written clearly. The results are mostly presented concisely. There are some paragraphs which are not written carefully and which should be strongly improved.
I have a few comments to the scientific content:
1. The authors use two methods to derive the ozone changes due to the particle impact, comparing the time series after the event to a) a lowess filtered time-series and b) 5-day average before the event assumed to be a quiet period. The authors completely dismiss a comparison between the two methods, and just from the inspections of their figures, they seem to show quite a different behaviour. At the end, the autors prefer to draw their conclusions from the second approach (Fig. 4). The authors should discuss the different approaches.
2. Inspecting Fig. 4 deltaO3, I am a bit puzzled by the persisting vertical structure of the signal seen at 80km (positive changes above, negative changes below). As this is just at the minimum between the secondary and tertiary ozone layer, a small vertical shift compared to the reference could also explain that pattern and the rather high values of the ozone change as the small reference goes into the denominator. So I doubt if the derived values are really significant. Absolute values would also clearer show the impact in the tertiary layer at 75 km.
3. Fig. 4 bottom shows the temperature change in the SH. First, showing percentage change for the temperature is not really meaningful in my opinion and I would ask you to show the absolute changes. Perhaps I misunderstood, but the explanation you give for the apparent temperature increase seems to me wrong: as the radiative damping time at these altitude is a few days only, any temperature signal caused by diabatic heating would be lost after some days. Secondly, the additional NO would rather increase the cooling rate, at least at the top level. So, probably changes in the dynamics, i.e. a descent of air masses is the reaseon for the temperature increase. But it is essentialy impossible to differ between adiabatic heating related to start of the seasonal descent and dynamical changes related to the event without specific experiments with a model of the MLT region. From Jia et al., dynamical changes related to elevated Ap-levels could be expected at about 100km and not at 80 km, contrary to your analysis.
4. The main finding seems to be that a major geomagnetic storm of this strength does not automatically lead to significant changes of ozone in the stratosphere. The discusssion in the conclusion paragraph is in my opinion a bit unclear. Specifically, as for the EPP indirect effect NOy transported from the mesosphere into the stratosphere is important, at least a discussion of observations of NO (eg. ACE/FTS) for this event would be helpful.
Minor comments:
First I have two general remarks: Many of the references the authors give do not point to the original papers, in some cases reviews or papers based on the original literature. So please check the references in this respect. Some typesetting (no italics for chemical substances, braces for citations) must be corrected, as you should a consistent tempus.
L2 SEP: you are using the abbreviation SEP here as solar energetic proton event, in L18 as Solar Energetic particle event, and in L112 just as solar energetic particles. Please be consistent.
L4 AURA/MLS instrument
L6 and at other places: O3 no italics
L6 upper part: perhaps MLT?
L12 please have a consistent tempus in this paragraph
L15 large perturbations : reference
L16 deacrease
L16 reference here and at many othe places in braces
L17 coronal mass ejections (CMEs)
L21 references
L22 typo Dst
L27 latitudes
L33 NOx is?
L34 define lower thermosphere
L35 Those species: sentence is wrong. Only NOx is long-lived in the winter mesosphere, and in the stratosphere only in the reservoire gases as NOy. HOx is shortlived everywhere.
L35 In the presence ...: not correct, as in the MLT region there is no polar vortex as its is found in the stratosphere. Here elevated stratopause events are important.
L41 Thus ... not generally true (self healing effct). Generally, many references in this paragraph are erratic as the presentation of the processes. Please rewrite.
L60 reference
L63 coverage is only to 82° (L60)?
L81 is this your strategy?
L86 reference to the lowess method? exact parameters?
L106 completely removed?
L112 Delete or rewrite this paragraph. How much ozone change do you expect from the particle flux and energy in the middle stratosphere?
Fig.1 improve color coding for less barrier
L127 because .... delete. Depends on the lifetime of the active region.
Fig.3 lines are hard to differ.
L155 computed with O3 vmr???
Fig.4 SH?
L156 whats some ozone?
L178 Joule
L177 please show that quantitativly
L178 here EPP is EE precipitation?
L179 energetic particles loose their energy by inelastic collisions (causing partly ionization), ending up in some heat, excitation, radiative loss. Joule heating is caused by ion drag and not directly related to the EPP.
L183 slow ozone depletion??
L209 is possible?
L222 This paragraph is not written carefully and somwhat speculative whithout enlightening. What is a quick response, what is hard enough, and how much ionization do you need to see an effect, how do you define upper stratosphere (60km??), etc.. Yes, the season is not favourable for an impact in the SH stratosphere because of the too short lifetime of NOx in the still sunlit mesosphere. Further analysis, for example regarding the descent, needs more observation of long-lived tracers and very specifically of NOy, or just applyication of an atmospheric chemistry model like WACCM.
Citation: https://doi.org/10.5194/angeo-2024-29-RC1 -
AC1: 'Reply on RC1', Alexandre Winant, 19 Feb 2025
Review of "Ozone decrease observed in the upper atmosphere following the May 11th 2024 Mother’s day solar storm" by A. Winant et al.
The authors analyse MLS observation in early 2025 for examining the impact of geomagnetic storm on May-11 2024 on ozone in the middle atmosphere and MLT region. They find additional ozone depletion related to the event for the secondary and tertiary ozone layer in the Southern hemisphere. In addition, they study the solar proton event in June 2024 for comparison. None of these events caused long-lived ozone depletion in the stratosphere.
Through its worlwide observable auroras, the geomagnetic storms of May 2024 draw broad public attention. But also from a scientific perspective, the geomagnetic storm in May 2024 is interesting as it exhibited the second strongest Ap-index recorded, and ranks among the strongest storms in terms of the Dst-index. The paper is one of the first publications to study specifically the impact of this storm on the neutral atmosphere. The paper is timely and of general interest.
The paper is generally written clearly. The results are mostly presented concisely. There are some paragraphs which are not written carefully and which should be strongly improved.
Thank you for your in-depth reading and your comments that helped us to improve the paper. The corrections in the text are in red for the removed sentences and in blue for the added sentences. Our answers are given below.
I have a few comments to the scientific content:
- The authors use two methods to derive the ozone changes due to the particle impact, comparing the time series after the event to a) a lowess filtered time-series and b) 5-day average before the event assumed to be a quiet period. The authors completely dismiss a comparison between the two methods, and just from the inspections of their figures, they seem to show quite a different behaviour. At the end, the autors prefer to draw their conclusions from the second approach (Fig. 4). The authors should discuss the different approaches.
We have added more explanations relate the results from the two methods both in the results as well as in the discussion and conclusion.
- Inspecting Fig. 4 deltaO3, I am a bit puzzled by the persisting vertical structure of the signal seen at 80km (positive changes above, negative changes below). As this is just at the minimum between the secondary and tertiary ozone layer, a small vertical shift compared to the reference could also explain that pattern and the rather high values of the ozone change as the small reference goes into the denominator. So I doubt if the derived values are really significant. Absolute values would also clearer show the impact in the tertiary layer at 75 km.
We have added the graphs with the absolute values in the figure for the May and June events. Moreover, for the large relative differences observed at 80 km, a sentence has been added to the text mentioning the effect that small values could have on the relative difference. In addition, the description of the observations in the secondary layer has been changed to highlight the observation at 84 km, where the relative difference is positive. - Fig. 4 bottom shows the temperature change in the SH. First, showing percentage change for the temperature is not really meaningful in my opinion and I would ask you to show the absolute changes. Perhaps I misunderstood, but the explanation you give for the apparent temperature increase seems to me wrong: as the radiative damping time at these altitude is a few days only, any temperature signal caused by diabatic heating would be lost after some days. Secondly, the additional NO would rather increase the cooling rate, at least at the top level. So, probably changes in the dynamics, i.e. a descent of air masses is the reaseon for the temperature increase. But it is essentialy impossible to differ between adiabatic heating related to start of the seasonal descent and dynamical changes related to the event without specific experiments with a model of the MLT region. From Jia et al., dynamical changes related to elevated Ap-levels could be expected at about 100km and not at 80 km, contrary to your analysis.
As suggested, we added panels with the absolute temperature profiles to the figures. Secondly, we removed the sentence that was referring to the processes involved in the heating that we observed in May as it was to speculative. In the last section, we did not want to us the results from Jia et al. to explain the decrease of ozone between 70 km and 80 km, but rather in the secondary layer above 90 km. We have changed the text to make it more clear. - The main finding seems to be that a major geomagnetic storm of this strength does not automatically lead to significant changes of ozone in the stratosphere. The discusssion in the conclusion paragraph is in my opinion a bit unclear. Specifically, as for the EPP indirect effect NOy transported from the mesosphere into the stratosphere is important, at least a discussion of observations of NO (eg. ACE/FTS) for this event would be helpful.
We agree that observations of NO would be interesting for the discussion but the addition of new data from a instrument with which we have not worked yet is difficult to do within the given time. We have reformulated this paragraph to make it less speculative for the involvement of NOx in the stratosphere for this particular event.
Minor comments:
First I have two general remarks: Many of the references the authors give do not point to the original papers, in some cases reviews or papers based on the original literature. So please check the references in this respect. Some typesetting (no italics for chemical substances, braces for citations) must be corrected, as you should a consistent tempus.
L2 SEP: you are using the abbreviation SEP here as solar energetic proton event, in L18 as Solar Energetic particle event, and in L112 just as solar energetic particles. Please be consistent.
Indeed the labeling was not consistent, this has been changed. Every occurrence is now labeled as ‘Solar Energetic Particle’ event.L4 AURA/MLS instrument :Added
L6 and at other places: O3 no italics: We changed for all occurrences of O3 and other chemical species.
L6 upper part: perhaps MLT? : Changed
L12 please have a consistent tempus in this paragraph: The tempus was changed to be more consistent.
L15 large perturbations : reference : the sentence before was slightly modified and a reference was added.
L16 deacrease : Corrected
L16 reference here and at many othe places in braces: Braces added to all references.
L17 coronal mass ejections (CMEs): The typo was corrected
L21 references : References were added to the text.
L22 typo Dst : We corrected the typo
L27 latitudes : corrected
L33 NOx is? : changed to Nitrogen oxides NOx are
L34 define lower thermosphere: we added (90 to 100 km)
L35 Those species: sentence is wrong. Only NOx is long-lived in the winter mesosphere, and in the stratosphere only in the reservoire gases as NOy. HOx is shortlived everywhere.
L35 In the presence ...: not correct, as in the MLT region there is no polar vortex as its is found in the stratosphere. Here elevated stratopause events are important.
L41 Thus ... not generally true (self healing effct). Generally, many references in this paragraph are erratic as the presentation of the processes. Please rewrite.Answer for L33, 34, 35, 41: As asked by the reviewer, we heavily redacted this paragraph of the introduction for more clarity and added references to original papers.
L60 reference : We added the reference: Schwartz, M.: MLS/Aura Level 2 Ozone (O3) Mixing Ratio V005, https://doi.org/10.5067/AURA/MLS/DATA2516, 2021
L63 coverage is only to 82° (L60)? : Yes, thank you for bringing this to our attention, the orbit of the MLS satellite limits its observations to 82° of latitudes in each hemisphere. We changed the 90° in the text to 82°. The 90° comes from the latitude filtering used in the code that is set to 90°.
L81 is this your strategy?: This is the approach we use. Changed in the text.
L86 reference to the lowess method? exact parameters? : The reference to the original paper was added with the value of the ‘frac’ (f = 0.25) parameter used to smooth the data.
L106 completely removed? : This was changed into : ‘depleted from around 2.5 ppmv to below 1 ppmv’
L112 Delete or rewrite this paragraph. How much ozone change do you expect from the particle flux and energy in the middle stratosphere? As suggested, we deleted this paragraph from the text.
Fig.1 improve color coding for less barrier : Figure 1 was changed to reduce the visibility of the grid. This was done for all images.
L127 because .... delete. Depends on the lifetime of the active region : The sentence was slightly changed, and we have added a reference that shows that the region that caused the SEP of May was still active in June after it had done one full rotation around the Sun.
Fig.3 lines are hard to differ. : The plot was updated with higher contrast to help better differ the different lines.
L155 computed with O3 vmr??? : As we have changed the figure that text was also changed.
Fig.4 SH? : Indeed, it is in the southern hemisphere. We added to the caption of Fig.4 and Fig.5
L156 whats some ozone? This sentence was removed since it was too vague and the exact ozone variations are described below.
L178 here EPP is EE precipitation? : Here we mean both protons and electrons. Changes to energetic particle precipitation in the text.
L177 please show that quantitatively
L178 Joule : Done
L179 energetic particles loose their energy by inelastic collisions (causing partly ionization), ending up in some heat, excitation, radiative loss. Joule heating is caused by ion drag and not directly related to the EPP.Answer to L177, L178, L179: This has to do with the scientific comment #3. This part of the text was revised.
L183 slow ozone depletion??: We clarified this sentence in the text.
L209 is possible?: We corrected the tempus
L222 This paragraph is not written carefully and somwhat speculative whithout enlightening. What is a quick response, what is hard enough, and how much ionization do you need to see an effect, how do you define upper stratosphere (60km??), etc.. Yes, the season is not favourable for an impact in the SH stratosphere because of the too short lifetime of NOx in the still sunlit mesosphere. Further analysis, for example regarding the descent, needs more observation of long-lived tracers and very specifically of NOy, or just applyication of an atmospheric chemistry model like WACCM.
We strongly reformulated this paragraph to remove the speculations on the effects of NOx for the events in question and keep our conclusions to the observations.
-
AC1: 'Reply on RC1', Alexandre Winant, 19 Feb 2025
-
RC2: 'Comment on angeo-2024-29', Anonymous Referee #2, 25 Feb 2025
This paper uses MLS observations to investigate the response of ozone in the polar regions of the middle atmosphere to two solar storms that occurred in May and June of 2024. For both storms, it finds no response in the northern hemisphere. For the May storm a large relative decrease around 75 km in the southern hemisphere is detected, but not in June. The reason given is that the total flux of energetic particles is different. The authors suggest a small depletion about 90 km after May 10 in the SH is from a temperature increase rather than directly related to EPP at those heights and NOx production.
While the paper contains some new results, I do not think authors have reached substantial conclusions. They do not calculate the total ion pair production rates from these storms and show that the May storm is different from the June storm or that EPP occurred at a significant level in the stratosphere. They do not provide an explanation for the hemispheric asymmetry in the response except to state the seasons are different. In addition, Figure 4 shows a larger increase (+30%) in ozone at 85 km than the discussed depletion (-10%) at 95 km, but it is not mentioned at all in the paper.
The authors should endeavour to better connect the different ozone responses to the different storm characteristics, provide a more complete explanation of the hemispheric asymmetry and distinguish between the chemical and dynamical impacts on ozone. It is certainly unclear to me if the 10% reductions in ozone above 90 km are caused by EPP or simply dynamical variations (random variability or the seasonal cycle). The paper really needs to show that the differences are statistically significant - perhaps comparisons to the previous May/June period would support the position that ozone is in fact responding to the particle inputs. Below are minor comments that it would be good to address.
Title: The title does not reflect that this is a comparison of two storms.
Abstract:
The term "solar energetic proton (SEP) event" seems to mix up solar energetic particle (SEP) and solar proton event (SPE). I think he latter term is more appropriate here.
"and reaches a maximum of 60%" - it would be helpful to say what altitude.
Make it clear that the signals are only present in the southern hemisphere.
Introduction:
"predicted in 2025" - a reference would be good to add here. In fact, it's generally accepted that the peak in storms occurs on the declining phase.
Correct the referencing style for inline reference - put them in parens.
Be consistent in your definition of SEP - is it "solar energetic proton" (line 2) or "Solar Energetic Particle" line 18, with the definition repeated on 112.
Duplicate definitions of EPP on 31 and 44.
Suggest: "Thus, *a* result of EPP is a decrease in"
"However, strong evidence of SEP directly depleting stratospheric ozone are scarce" - I don't think this is true. See doi:10.5194/acp-11-9089-2011 where ozone is depleted immediately during the SPE down to 10 hPa (i.e., into the upper startosphere. Also the many papers led by Charlie Jackman (e.g., https://acd-ext.gsfc.nasa.gov/People/Jackman/Jackman_2004.pdf). The evidence is strong, but large SPEs themselves are rare.
Data:
Please include some discussion of the vertical resolution of the MLS data in the stratosphere and MLT. Also, averages are presented from 60-90˚ Does MLS measure all latitudes uniformly, or is a simple average of all profiles above 60˚? If the latter, does the sampling change over the period being analysed and what changes might that create? Also, make it clear you are using MLS temperatures - we only see them in the analysis in Figure 4.
Results:
What do you mean by "but not always" - this is vague. Please state when, in this 6 month period, you have an SEP and no evidence of ozone depletion.
In regard to Figure 1 the authors state "Those variations on smaller time scales are not linked to geomagnetic storms illustrated by high peaks of geomagnetic activity in the bottom panel, in any of the ozone layers." To me it looks like they are. Especially in the SH, and stated by the authors on line 120.
"because they originate from the same region" - Can you provide some more evidence that the same active region erupted to create these events.
Figure 3 is just too small to distinguish the variations. Please make the Y scale much larger. The colours of the lines do not match the legend.
Can you provide evidence that the temperature variation is not just seasonal dynamical variability? Perhaps show 2022 and 2023 for comparison?
149/Figure 4: Is this NH or SH? Presumably the latter but it's not clear.
214: "In the MLT region...unlike in the mesosphere ..." Since the M in MLT is mesosphere this statement does not make sense. Better to state the altitude ranges you are talking about explicitly.
223: "In both cases, the spectrum of solar protons was hard enough to produce ionization in the upper stratosphere." This has not been shown and does not describe exactly how much might have been produced?
225: "This absence of response in stratospheric ozone could be explained by the season again." How, and why would this impact both hemispheres? i.e., both in winter and summer.
Conclusion:"In the MLT region... the decrease in ozone are not linked to catalytic reactions with HOx and N Ox"
If energetic electrons precipitated, they would produce HOx in the MLT (the MLT includes the region below the mesopause) and would therefore destroy ozone.No evidence has been provided that there was a change in the circulation or that it impacted O and H transport.
Figure 1 caption - "ration" typo
Citation: https://doi.org/10.5194/angeo-2024-29-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 511 | 76 | 23 | 610 | 8 | 12 |
- HTML: 511
- PDF: 76
- XML: 23
- Total: 610
- BibTeX: 8
- EndNote: 12
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1